pip install -r requirements.txtThe command to run benchmark:
bash run_benchmark.sh ${inference_dir}About the above command, the The directory ${inference_model_dir} should contain the Paddle Inference model files:
inference_std
├── model_list.txt
├── PPLCNet_x0_25
│ ├── inference.pdiparams
│ ├── inference.pdiparams.info
│ └── inference.pdmodel
├── PPLCNet_x0_35
│ ├── inference.pdiparams
│ ├── inference.pdiparams.info
│ └── inference.pdmodel
├── PPLCNet_x0_5
│ ├── inference.pdiparams
│ ├── inference.pdiparams.info
│ └── inference.pdmodel
└── PPLCNet_x0_75
├── inference.pdiparams
├── inference.pdiparams.info
└── inference.pdmodel
And the model_list.txt should describe the models to be tested:
PPLCNet_x0_25
PPLCNet_x0_35
PPLCNet_x0_5
PPLCNet_x0_75
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent